Journal Pre-proof

Reaction kinetics of supplementary cementitious materials in

reactivity tests

Sivakumar Ramanathan , Luis Ruiz Pestana , Prannoy Suraneni

PII: S2666-5492(22)00002-0
DOI: https://doi.org/10.1016/j.cement.2022.100022
Reference: CEMENT 100022

To appear in: Cement

Received date: 28 October 2021

Revised date: 22 December 2021
Accepted date: 25 February 2022

Please cite this article as: Sivakumar Ramanathan , Luis Ruiz Pestana , Prannoy Suraneni , Re-
action kinetics of supplementary cementitious materials in reactivity tests, Cement (2022), doi:
https://doi.org/10.1016/j.cement.2022.100022

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.

© 2022 Published by Elsevier B.V.

This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/)
 Reaction kinetics of supplementary cementitious materials in reactivity tests

Sivakumar Ramanathan1, Luis Ruiz Pestana2, Prannoy Suraneni2,*

1
Civil and Construction Engineering, Oregon State University, Corvallis, OR 97331, USA
2
Civil and Architectural Engineering, University of Miami, Coral Gables FL 33146

*Corresponding author: Prannoy Suraneni (suranenip@miami.edu)

Abstract

This work characterizes the reaction kinetics of supplementary cementitious materials

(SCMs) with calcium hydroxide in the modified R3 test. The heat flow curves of 58 SCMs of

varying reactivities were studied. Based on the heat flow curves, the SCMs were classified as

more reactive, less reactive, and inert. Most of the heat flow curves in the modified R3 test

exhibit, after the peak of heat flow, an initial slower decaying power-law regime that transitions

into a longer and faster decaying power-law regime. The pre-exponent of the first regime

depends on the initial SCM reactivity and correlates well with the 24-hour heat release in the

modified R3 test, thus making it a useful metric for rapid classification of SCMs.

Keywords: Supplementary cementitious materials; reactivity; kinetics.

1. Introduction

The use of supplementary cementitious materials (SCMs) as partial replacement to

ordinary portland cement (OPC) in concrete reduces CO2 emissions associated with cement

manufacture and generally improves concrete durability [1-4]. The availability of conventional

SCMs, such as fly ash, has decreased due to the phasing out of the relevant industries (e.g., coal
power plants) because of environmental and cost concerns. Therefore, there is a pressing need to

use alternative or novel SCMs in concrete. However, to effectively use novel SCMs, their

reactivity needs to be assessed and understood. Reactivity tests such as the strength activity

index test, the Chapelle test, or the Frattini test have been used for years for conventional and

alternative SCMs but have critical flaws [5-8]. As an example, for the strength activity index test,

the variable water-to-cementitious materials ratio and the early age of testing obscure the

contribution of the SCM reaction, and consequently lead to potential misclassification of inert

materials [8]. The Chapelle and Frattini tests are based on portlandite consumption, and thus are

not truly applicable to latent hydraulic materials such as slag [5-7].

Recent advances in reactivity testing have yielded four promising SCM reactivity tests –

the bulk resistivity index test [8], modified lime strength test [9], the R3 test [10], and the

modified R3 test [11], which overcome most of the shortcomings of previous tests. Therefore,

these tests can be used to classify SCMs based on their reactivity [12]. The R3 test measures 7-

day heat release and/or bound water content in model systems containing SCM and calcium

hydroxide in a high-pH solution at 40 °C with added sulfates and carbonates [10,13]. The R3 test

is reflective of 28-day behavior, as it is based on comparisons with 28-day strength data [13].

However, it may not be well-suited for slowly reacting materials such as Class F fly ash and

other siliceous materials, which show limited extent of reactivity at 7 and 28 days [8]. In

addition, using heat release or bound water does not easily allow the differentiation of latent

hydraulic and pozzolanic materials, which may show similar values of heat release while having

very different effects in concrete [11]. The modified R3 test [11] measures the 10-day heat

release and calcium hydroxide consumption in similar model systems at 50 °C but without added

sulfates and carbonates. Removing the extra sulfates and carbonates simplifies the system. The
exact roles of sulfates and carbonates in the reaction are complex and vary depending on whether

the SCMs are largely siliceous, aluminosilicate, or calcium aluminosilicate in terms of their

composition [14]. The higher temperature and longer duration make this test more suitable for

slowly reacting materials. Since this test is highly accelerated, its outputs are typically reflective

of the behavior of SCMs in cement pastes (or concrete) at later ages (beyond 56 days) [15,16].

By measuring both heat release and calcium hydroxide consumption, the modified R3 test can be

used to classify SCMs into inert (IN), pozzolanic, less reactive (P, LR), pozzolanic, more

reactive (P, MR), latent hydraulic, less reactive (LH, LR), and latent hydraulic, more reactive

(LH, MR) categories [12]. This ability for classification is a critical advantage as no other test

has this ability.

Early-age measurements of heat release in reactivity tests, which could provide

important insight into the behavior of cementitious mixtures, have not been fully explored.

Isothermal calorimetry offers a continuous measurement of heat release, however, only the

ultimate heat release has commonly been used as a measure of reactivity. Therefore,

understanding the reaction kinetics in the modified R3 test and other reactivity tests is critical to

obtain a full picture of SCM reactivity and to correlate SCM behavior in model systems and

cementitious paste properties [17].

Analytical models used to describe the hydration kinetics of cementitious systems, such

as C3S mixed with water [18], generally rely on Avrami-type nucleation and growth formulations

which assume an exponential or stretched-exponential decay in the heat flow after the peak [19].

Power law fits, arising from the complex interplay of dissolution, nucleation and growth, and

diffusion phenomena, have also been suggested [18]. More sophisticated numerical methods,

such as thermodynamic modeling or the Parrot and Killoh model, have also been used to study
hydration of OPC and SCM-OPC/C3S systems [20-26]. Other methods like ab initio calculations,

kinetic Monte Carlo simulations or mathematical models of dissolution, nucleation, and growth,

have also been used to investigate different aspects of the hydration process [18,19,27-29]. The

reaction kinetics of pure SCM model systems (with calcium hydroxide at high pH), however,

remain poorly understood, and it is unclear whether classical hydration kinetic models apply.

In this study, for the first time, the reaction kinetics in the modified R3 test are

characterized and modeled. Specifically, this is done using 58 SCMs with varying reactivities

run for test a duration of 10 days. The potential of the heat flow curves and fit parameters to be

used for rapid SCM classification is evaluated.

2. Materials and Methods

A total of 58 SCMs of varying levels of reactivity and different classifications – inert,

pozzolanic, and latent hydraulic, were evaluated in this study. The SCMs include 4 basalt fines

(BF1-4), 7 biomass ashes (BMA1-7), 4 calcined clays (CC1-4), 12 Class C fly ashes (CFA1-12),

12 Class F fly ashes (FFA1-12), 2 glass powder pozzolans (GP1-2), 2 silica fumes (SF1-2), 5

ground granulated blast furnace slags (SL1-5), 2 pumices (P1-2), 2 limestone powders (LS1-2), 1

each of nepheline syenite filler (NF), ground bottom ash (GBA), sandstone powder (SST), quartz

(Q), ladle furnace slag (LFS), and a processed silica SCM (SP). The chemical compositions of

these SCMs (from x-ray fluorescence) and d50 values (from laser diffraction), where available,

are provided in Table A1 (Supplementary Material). More details regarding BF1 can be found in

[12], BMA6-7 in [30,31], MK1-3 in [12,32], GP1-2 in [12,15], SF1-2 in [15,16], SL1-5 in [12],

P1-2 in [15], LS1-2 in [16], LFS in [33], and, BMA1-5, and NF, GBA, SST, and SP in [34].
While reactivity data for some materials has been published, reaction kinetics and their modeling

have not been investigated.

2.1. Modified R3 test

The modified R3 test was carried out on all SCMs based on the method described in

Suraneni and Weiss [11]. Reagent grade calcium hydroxide and the SCM (mass ratio 3:1) were

dry mixed together for four minutes in a plastic container to ensure uniform distribution. A

freshly prepared 0.5 M potassium hydroxide solution was added to the dry powder mixture at a

liquid-to-solid ratio of 0.9. Because of the small quantities of components involved, they were

hand mixed thoroughly for four minutes. From this mixture, 6-7 g were placed in a glass

ampoule, sealed, and put into an isothermal calorimeter (TAM Air, TA Instruments)

preconditioned at 50 ± 0.05 °C. The heat flow data for the first 45 minutes was not collected due

to temperature differentials. The heat release parameters were collected for 10 days. At the end

of this period, thermogravimetric analysis (TGA55, TA Instruments) was performed using

approximately 30-50 mg of the sample obtained from the bulk. The temperature was ramped at

10 °C/minute from ambient to 600 °C in an inert nitrogen atmosphere. The mass loss in the range

of approximately 380-460 °C was measured using the tangential method described in Kim and

Olek [35] and used to determine calcium hydroxide content, from which calcium hydroxide

consumption was calculated. For selected specimens, TGA was carried out up to 1000 °C to

check for carbonation; significant carbonation was not detected. TGA was performed within 12

hours of completing the isothermal calorimetry measurements and within 15 minutes of breaking

the ampoule. Replicate testing of five randomly chosen SCMs showed that the differences in

heat release and calcium hydroxide consumption values between specimens were less than 5%.

The SCMs were classified using previously established criteria [12,16]. The classification results
are shown in Fig. A1 (Supplementary Material). Based on the classification, BF, LS, NF, SST,

LFS and Q were classified as inert; P, GP, BMA1, BMA6-7, FFA, CFA were classified as

pozzolanic, less reactive; SF, CC, BMA2-5, SP, and GBA were classified as pozzolanic, more

reactive; SL was classified as latent hydraulic, more reactive. Five of the 58 SCMs could not be

unequivocally assigned to any classes, which highlights a limitation of the classification criteria.

Nonetheless, those SCMs were assigned to the class to which their distance to the classification

boundary was the least.

3. Results and Discussion

Fig. 1 shows the heat release curves of example Class C and Class F fly ashes. Up to 3

days, the Class C fly ash was significantly more reactive than the Class F fly ash, with heat

release values being up to 50% higher. The Class F fly ash reacted in a slower but more

sustained manner, and at the end of the test (10 days), the difference in the heat release values of

the two materials was insignificant (5%).

 Figure 1. Heat release curves of example Class C (CFA3) and Class F (FFA2) fly ashes.

Numerous Class C and Class F fly ashes have been studied in our group using the

modified R3 test [12], and this behavior always appeared in the heat release response. The

reactivity of the fly ashes in the modified R3 test was consistent with their reactivity in

cementitious pastes (albeit at different time scales due to the accelerated nature of the modified

R3 test). These differences in reactivity are driven by differences in the reactivities of the

siliceous, aluminosilicate, and calcium aluminosilicate glasses that comprise the fly ashes [14]. It

is critical to understand the reaction kinetics; otherwise, incorrect assessments of material

reactivity could be made.

3.1. Modeling the heat flow curves and prototypical heat flow curves for different SCM classes

The data for the first 45 minutes to one hour after mixing in the modified R3 test were not

collected due to temperature differentials. Generally, the first peak in the rate of heat release
occurred in this time interval and thus was not measured for most of the SCMs studied.

However, for nine slower reacting SCMs, the heat flow before and leading to the peak value was

captured. Because here we only focus on modeling the post-peak evolution of the heat flow (i.e.,

the derivative of the heat release), the aforementioned nine SCMs were not modeled for

consistency.

The reaction of the SCMs and calcium hydroxide in the modified R 3 test is a complex

process and the reaction rate depends on factors including but not limited to the glass content, the

chemical composition of the calcium aluminosilicate glass, activation energy, and SCM fineness

[36]. The 240-hour heat flow curves of most of the non-inert SCMs studied here could be best

captured by two power-law regimes (Fig. 2): a shorter, slower decaying regime (Eq. 1) that

transitioned at time tth into another longer, faster decaying one (Eq. 2). The reaction kinetics

according to this model were described by four parameters A, B, and α, β which correspond to

the pre-exponential factors (i.e., the intercepts at t = 1 hour) and the exponents (i.e., the slope in a

log-log scale), of each of the regimes respectively:

Q = At-α when t < tth (1)

Q = Bt-β when t > tth (2)

where Q and t are heat flow (in mW/g SCM) and time (in hours).
 Figure 2. Fitting procedure and calculation of threshold time.

The pre-exponent A is determined as the intercept with the y-axis at t = 1 hour, and is

thus an indicator of the early heat flow. The exponents α and β capture how fast the heat flow

decayed over time, with larger values implying faster decays. To calculate the parameters of the

model, the initial part of the curve (0-4 hours) and the tail of the curve (61-240 hours) were fit to

the respective power-law regimes. The threshold time, tth (which is not a fitting parameter), was

determined as the time where the two power-laws intersected. The values of the parameters were

not sensitive to the range of time scales used to fit the different regimes. For example, the

differences in the pre-exponent values were less than 15% when using 0-6 hours and 39-240

hours to fit the first and second regimes respectively.

The heat flow curves from the modified R3 test generally fell into three prototypical curve

shapes, which are referred to as type MR (more reactive), LR (less reactive), and IN (inert) (Fig.
3). In 16 out of the 58 SCMs studied here, the tth values were either less than t = 1 hour or more

than t = 240 hours. In such cases, mostly corresponding to IN or MR curves, the heat flow was

modeled using a single power-law (Q = At-α).

For MR SCMs, the parameter A, related to the initial heat flow, was between 10-40

mW/g SCM. The heat release curves for the MR SCMs showed a rapid initial increase in the heat

release, with a slower increase at later ages. SCMs such as silica fume, metakaolin/calcined

clays, Class C fly ashes, and blast furnace slags followed this type of curve. These materials are

either aluminosilicates, calcium aluminosilicates, or fine siliceous materials.

LR SCMs had a much lower initial heat flow (A in the range of 1-5 mW/g SCM), which

decayed extremely slowly for about 20-100 hours (the tth), and then more rapidly, ultimately

reaching a similar value as the MR SCMs. At early times, the heat flow of the MR SCMs was

higher than for LR SCMs; but at later ages, there was a crossover. In the heat release plot, Type

LR curves showed a gradual increase in the heat release for 20-100 hours, but the rate of change

decreased with time. Examples of LR SCMs include Class F fly ashes, pumice, and glass powder

– all siliceous materials. The siliceous materials do not necessarily show low values of heat

release (Fig. 1) and are perhaps better termed as slowly reactive materials [14].

IN curves had low initial heat flow values (1 mW/g SCM or below), and typically

exhibited a single power-law regime. A sharp decrease in the heat flow at around 200 hours was

seen in IN curves, although it was unclear if this drop was an experimental artefact due to the

low heat flow values which made the response indistinguishable from noise [11,12]. The final

heat flow values for IN SCMs were significantly lower than those of MR or LR SCMs.

Examples of IN SCMs include limestone and quartz. IN SCMs showed a low heat release
through the duration of the experiment. The fitting parameters for all SCMs in this work are

provided in Table A2 (Supplementary Material).

The prototypical curve shapes (MR, LR, and IN) were based solely on the curve shape

and provide important reaction kinetics information. On the other hand, the SCM classifications

(IN; P, LR; P, MR; LH, LR; and LH, MR) were based on the 10-day value of the heat release

and calcium hydroxide consumption and provide indicators of overall reaction and type of

reaction [12]. SCM classifications did not directly translate into kinetic information nor vice

versa.
 Figure 3. Typical curves for SCMs in the modified R3 test – (a) Heat release and (b) Heat flow

(log-log).

3.2. Model parameters and different SCM classes

The values of the fit parameters of the kinetic model were investigated for different SCM

classes. Fig. 4 shows for different SCM classes the values of the pre-exponent A, which is

related to the initial heat flow, the exponent α, which relates to how fast the heat flow decreases

over time, and the threshold time tth, at which the reaction kinetics transition from a slowly

decaying regime to fast decaying one. The range of these values for all SCM classes are shown

in Table A3 (Supplementary Material).

The value of A was one or two orders of magnitude greater for reactive materials than for

the inert materials. The fact that differences in SCM behavior could be identified using model

parameters that can be determined within the first few hours of testing is significant as test times
could be substantially reduced. While the test duration can be reduced to 1-day or 3-days, this

should not be done for the slowly reacting LR SCMs without careful kinetic corrections [14]. For

Class F fly ashes (μ = 5, σ = 5), A values were significantly lower than that of Class C fly ashes

(μ = 11, σ = 3), where μ and σ are the mean and the standard deviation, respectively.

The values of α did not strongly depend on the SCM class except for IN SCMs. For

Class F fly ashes (μ = 0.4, σ = 0.2), the α values were about 40% lesser than those of Class C fly

ashes (μ = 0.6, σ = 0.2). The range of values for tth for the different SCM classes somewhat

overlapped (Fig. 4c); however, there was a significant difference in the average values of P, LR

and P, MR SCMs. Threshold times could not be determined for some P, MR SCMs (such as

calcined clays and silica fume) due to their rapid reaction and for some P, LR SCMs, as the

convergence was beyond 240 hours. The average tth for Class F fly ashes (μ = 63 hours, σ = 42

hours) was more than twice that of Class C fly ashes (μ = 27 hours, σ = 33 hours). Our results

show that, initially, reaction rate for Class C fly ashes was greater than that for Class F fly ashes,

but decreased faster; at later ages, reaction rates for Class F fly ashes were higher [8]. This

finding highlights the importance of considering the entire heat curves and not just the reactivity

at 10 days, especially for slowly reacting materials. The behavior in these model systems was

consistent with the behavior seen in concrete for different fly ash classes [8,17,36]. Although

SCM behavior in a cementitious system depends on several factors [36], these results indicate a

qualitative estimate of reactivity could be obtained based on A, α, and tth.

Figure 4. Range of parameters based on power law model – (a) Pre-exponent A, (b) Exponent α,

and (c) Threshold times tth. The crosses on either end of the box plot represent the maximum and

minimum values, the top and bottom sides of the box represent the upper and lower quartile

values of the dataset, and the dot and line inside the box are the mean and median of the dataset.

Fig. 5 shows the relationships between pairs of model parameters, for the different SCM

classes shown in different colors. Fig. 5a shows that Class C fly ash (blue triangles) and Class F

fly ash (red dots) occupied distinct regions in the tth vs. A log-log plot. Even though both these

SCMs could be classified as P, LR, they follow completely different trends, because the P, LR

classification did not consider reaction kinetics [8]. The slags (pink diamonds) formed a distinct

cluster from other SCMs, likely because of their latent hydraulic properties [12]. As one moves

from left to right of the plot, the SCM reactivity increased in the direction represented by the

arrow. More reactive materials were characterized by high A values and earlier tth, whereas less
reactive materials had low values of A and later tth. However, for Class C fly ashes, there was no

correlation between tth and A.

The semi-log plot of A and α in Fig. 5b shows that there was a correlation between these

parameters when all SCMs are considered together. The reactivity of SCMs increased as we

move from the right to left of the plot in the direction indicated. Further, Class C fly ashes

exhibited intermediate reactivity, as observed in both Fig. 5a and Fig. 5b, where they fell

between the more and less reactive SCMs.

Fig. 6a shows the relationship between the 1-day heat release and the values of the

parameter A for all the SCMs. There was a moderate, positive correlation between the two

parameters. There was clear distinction between inert and reactive SCMs except for two

materials. In principle, the test duration could be reduced to one day if the objective were solely

to differentiate inert and reactive materials, which is an important practical consideration

[2,8,10]. Fig. 6b also shows the moderate correlation between the 3-day heat release and the A

value. The separation between inert and reactive SCMs was clearer, as the less reactive Class F

fly ashes reacted in the 1-to-3-day period. SCMs of different classes appeared somewhat

clustered, which was expected, since only heat release (and not calcium hydroxide) was being

used as a reactivity measure [11,12]. At 10 days, while the correlation between the parameters

was poorer, there was a clear distinction between reactive and inert SCMs. At all ages, reactivity

increased as we move from the right to left of the plot, with some outliers for Class F fly ash,

which showed low early age reactivity.

Figure 5. Two parameter plots for the variables from heat flow curves – (a) tth vs. A and (b) A vs.

α. FFA refers to Class F fly ash and CFA to Class C fly ash.
Figure 6. Two parameter plots of heat release and A value in the modified R3 test – (a) 1 day, (b)

3 days, and (c) 10 days.

3.3. Practical considerations

Results on a small set of materials suggest that the fit approach developed here is also

applicable for the R3 test [37]. Two important, practical findings are worth discussing. The first

is that by testing at 50 °C, reactive and inert materials could be differentiated as early as 1-day.

The demonstrated correlations between different direct and indirect measures of reactivity (heat

release, calcium hydroxide consumption, and bound water) from the R3 and modified R3 tests

[8,11,12,14,37] mean that other tests could also be shortened to 1-day regardless of whether their

exact reaction kinetics. While the tests could be shortened to 1-day, it does not appear that doing

so will provide a complete picture of the reactivity. Caution should be exercised for Class F fly

ashes and other slowly reacting siliceous materials, which show different reaction kinetics.
Indeed, research has shown that the 1-day or 3-day heat release in both the R3 and the modified

R3 test are strong predictors of the 10-day heat release, expect for siliceous SCMs [14].

The second important finding is the discovery that the reaction kinetics can be highly

variable, even in SCMs showing similar overall reactivity. When discussing reactivity test

results, parameters indicative of reaction kinetics should be presented along with those indicative

of overall reactivity to provide a better picture of SCM reactivity. A potential, simple candidate

to provide kinetic information would be the ratio of 3-day and 1-day heat release. This parameter

averaged 1.94 for the Class F fly ashes and 1.36 for the Class C fly ashes and generally the

highest values of this parameter were found for siliceous SCMs. Therefore, by stopping the test

at 3-days and using the 3-day and 1-day heat release ratio, a more complete picture of reactivity

and reaction kinetics could be rapidly obtained. Extrapolations and curve fitting the heat release

could also be used to estimate the 10-day or even ultimate heat release.

The complexity of the kinetics is important to consider when using the single point bound

water measurements, as is suggested in the R3 test (ASTM C1897). While the bound water

values will likely distinguish between inert and reactive materials, kinetic information cannot be

obtained from them. It could be advantageous to use bound water measurements at two times (1-

day and 3-days), especially for siliceous materials.

4. Conclusions

Reaction kinetics of SCMs in the modified R3 test were evaluated in this study. The

major conclusions from this study are –

1. Three prototypical curves were observed to describe the SCM behavior. These curves differ in

the initial value of heat flow and the initial and final slope of the heat flow.
2. The reaction in the modified R3 test can generally be captured by two power law regimes. The

parameters describing the power law fits, A, α, and tth, provide information about SCM reactivity

and classification. A is an especially powerful parameter, as it allows for rapid separation of

reactive and inert materials potentially within one day.

3. SCMs with similar overall reactivity can show widely varying reaction kinetics, meaning that

kinetic information must be considered when discussing SCM reactivity, a finding which has

important practical implications.

5. Acknowledgements

The work presented here was funded by the National Science Foundation (CMMI-

2101961). Partial support for the first author through the Graduate Dissertation Fellowship

(University of Miami) and for the last author from the Knight Foundation Endowment in the

College of Engineering (University of Miami) is gratefully acknowledged. Michael Croly and

Montale Tuen (University of Miami) are thanked for their help with the experimental work.

Declaration of interests

The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.

6. References
[1] B. Lothenbach, K. Scrivener, R.D. Hooton, Supplementary cementitious materials, Cem.
Concr. Res. 41 (2011) 1244-1256. https://doi.org/10.1016/j.cemconres.2010.12.001.
[2] R. Snellings, Assessing, understanding and unlocking supplementary cementitious materials,
RILEM Tech. Lett. 1 (2016) 50. https://doi.org/10.21809/rilemtechlett.2016.12.
[3] M. Juenger et al., Supplementary cementitious materials for concrete: Characterization needs,
MRS Proceedings (2012) 1488. https://doi.org/10.1557/opl.2012.1536.
[4] K. Scrivener, R.J. Kirkpatrick, Innovation in use and research on cementitious material, Cem.
Concr. Res. 38 (2008) 128-136. https://doi.org/10.1016/j.cemconres.2007.09.025.
[5] V.A. Quarcioni, F.F. Chotoli, A.C.V. Coelho, M.A. Cincotto, Indirect and direct Chapelle’s
methods for the determination of lime consumption in pozzolanic materials, IBRACON Struct.
Mater. J. 8 (2015) 1-7. https://doi.org/10.1590/S1983-41952015000100002.
[6] E. Ferraz, S. Andrejkovičová, W. Hajjaji, A.L. Velosa, A.S. Silva, F. Rocha, Pozzolanic
activity of metakaolins by the French standard of the modified Chapelle test: A direct
methodology, Acta Geodyn. Geomater. 12 (2015) 289-298.
https://doi.org/10.13168/AGG.2015.0026.
[7] J. Pontes, A. Santos Silva, P. Faria, Evaluation of pozzolanic reactivity of artificial
pozzolans, Mater. Sci. Forum 730 (2012) 433-438.
https://doi.org/10.4028/www.scientific.net/MSF.730-732.433.
[8] Y. Wang, L. Burris, C.R. Shearer, D. Hooton, P. Suraneni, Strength activity index and bulk
resistivity index modifications that differentiate inert and reactive materials, Cem. Concr.
Compos. 124 (2021) 104240. https://doi.org/10.1016/j.cemconcomp.2021.104240.
[9] M. Kasaniya, M.D.A. Thomas, E.G. Moffatt, Development of rapid and reliable pozzolanic
reactivity test method, ACI Mater. J. 116 (2019) 145-154. https://doi.org/10.14359/51716718.
[10] F. Avet, R. Snellings, A.A. Diaz, M.B. Haha, K. Scrivener, Development of a new rapid,
relevant and reliable (R3) test method to evaluate the pozzolanic reactivity of calcined kaolinitic
clays, Cem. Concr. Res. 85 (2016) 1-11. https://doi.org/10.1016/j.cemconres.2016.02.015.
[11] P. Suraneni, J. Weiss, Examining the pozzolanicity of supplementary cementitious materials
using isothermal calorimetry and thermogravimetric analysis, Cem. Concr. Compos. 83 (2017)
273-278. https://doi.org/10.1016/j.cemconcomp.2017.07.009.
[12] P. Suraneni, A. Hajibabaee, S. Ramanathan, Y. Wang, J. Weiss, New insights from
reactivity testing of supplementary cementitious materials, Cem. Concr. Compos. 103 (2019)
331-338.
https://doi.org/10.1016/j.cemconcomp.2019.05.017.
[13] X. Li, et al., Reactivity tests for supplementary cementitious materials: RILEM TC 267-
TRM phase 1, Mater. Struct. 51 (2018). https://doi.org/10.1617/s11527-018-1269-x.
[14] Y. Wang, S. Ramanathan, L. Burris, R.D. Hooton, C. Shearer, P. Suraneni, Reactivity of
unconventional fly ashes, SCMs, and fillers: Effects of sulfates, carbonates, and temperature,
Adv. Civ. Eng. Mater. (2021) (under review).
[15] S. Ramanathan, M. Kasaniya, M. Tuen, M.D.A. Thomas, P. Suraneni, Linking reactivity
test outputs to properties of cementitious pastes made with supplementary cementitious
materials, Cem. Concr. Compos. 114 (2020) 103742.
https://doi.org/10.1016/j.cemconcomp.2020.103742.
[16] S. Ramanathan, M. Croly, P. Suraneni, Comparison of the effects that supplementary
cementitious materials replacement levels have on cementitious paste properties, Cem. Concr.
Compos. 112 (2020) 103678. https://doi.org/10.1016/j.cemconcomp.2020.103678.
[17] M.D.A. Thomas, Optimizing the use of fly ash in concrete, Portland Cement Association
Vol. 5420, Skokie, Illinois, USA (2007).
[18] J.J. Thomas, A new approach to modeling the nucleation and growth kinetics of tricalcium
silicate hydration. J. Am. Ceram. Soc. 90 (2007) 3282-3288.
https://doi.org/10.1111/j.1551-2916.2007.01858.x.
[19] J.J. Thomas, et al., Modeling and simulation of cement hydration kinetics and
microstructure development, Cem. Concr. Res. 41 (2011) 1257-1278.
https://doi.org/10.1016/j.cemconres.2010.10.004.
[20] J. Lapeyre, T. Han, B. Wiles, H. Ma, J. Huang, G. Sant, A. Kumar, Machine learning
enables prompt prediction of hydration kinetics of multicomponent cementitious systems, Sci.
Rep. 11, (2021) 3922. https://doi.org/10.1038/s41598-021-83582-6.
[21] R. Snellings, G. Mertens, Ö. Cizer, J. Elsen, Early age hydration and pozzolanic reaction in
natural zeolite blended cements: Reaction kinetics and products by in situ synchrotron X-ray
powder diffraction, Cem. Concr. Res. 40 (2010) 1704-1713.
https://doi.org/10.1016/j.cemconres.2010.08.012.
[22] K. Scrivener, A. Nonat, Hydration of cementitious materials, present and future, Cem.
Concr. Res. 41 (2011) 651-665. https://doi.org/10.1016/j.cemconres.2011.03.026.
[23] J.J. Thomas, H.M. Jennings, J.J. Chen, Influence of nucleation seeding on the hydration
mechanisms of tricalcium silicate and cement, J. Phys. Chem. C 113 (2009) 4327-4334.
https://doi.org/10.1021/jp809811w.
[24] L.J. Parrot, D.C. Killoh, Prediction of cement hydration, Proc. Brit. Ceram. Soc. 35 (1984),
41-54.
[25] D. Glosser, P. Suraneni, O.B. Isgor, W.J. Weiss, Estimating reaction kinetics of
cementitious pastes containing fly ash, Cem. Concr. Compos. 112 (2020) 103655.
https://doi.org/10.1016/j.cemconcomp.2020.103655.
[26] J.J. Thomas, H.M. Jennings, Effects of D2O and mixing on the early hydration kinetics of
tricalcium silicate, Chem. Mater. 11(1999) 1907-1914. https://doi.org/10.1021/cm9900857.
[27] T. Oey, A. Kumar, J.W. Bullard, N. Neithalath, G. Sant, The filler effect: The influence of
filler content and surface area on cementitious reaction rates, J. Am. Ceram. Soc. 96 (2013)
1978-1990. https://doi.org/10.1111/jace.12264.
[28] J. W. Bullard, G. W. Scherer, J.J. Thomas, Time dependent driving forces and the kinetics
of tricalcium silicate hydration, Cem. Concr. Res. 74 (2015) 26-34.
https://doi.org/10.1016/j.cemconres.2015.03.016.
[29] K.A. Riding, J.L. Poole, K.J. Folliard, M.C.G. Juenger, A.K. Schindler, Modeling hydration
of cementitious systems, ACI Mater. J. 109 (2012) 225-234.
[30] M. Shakouri, C.L. Exstrom, S. Ramanathan, P. Suraneni, Hydration, strength, and durability
of cementitious materials incorporating untreated corn cob ash, Constr. Build. Mater. 243 (2020)
118171. https://doi.org/10.1016/j.conbuildmat.2020.118171.
[31] M. Shakouri, C.L. Exstrom, S. Ramanathan, P. Suraneni, J.S Vaux, Pretreatment of corn
stover ash to improve its effectiveness as a supplementary cementitious material in concrete,
Cem. Concr. Compos. 112 (2020) 103658. https://doi.org/10.1016/j.cemconcomp.2020.103658.
[32] S. Ramanathan, H. Moon, M. Croly, C.W. Chung, P. Suraneni, Predicting the degree
of reaction of supplementary cementitious materials in cementitious pastes using a pozzolanic
test, Constr. Build. Mater. 204 (2019) 621-630.
https://doi.org/10.1016/j.conbuildmat.2019.01.173.
[33] Y. Wang, P. Suraneni, Experimental methods to determine the feasibility of steel slags as
supplementary cementitious materials, Constr. Build. Mater. 204 (2019) 458-467.
https://doi.org/10.1016/j.conbuildmat.2019.01.196.
[34] S. Ramanathan et al., Reactivity of alternative supplementary cementitious materials, Cem.
Concr. Res. 2021 (under review).
[35] T. Kim, J. Olek, Effects of sample preparation and interpretation of thermogravimetric
curves on calcium hydroxide in hydrated pastes and mortars, Transp. Res. Rec. 2290 (2012) 10-
18. https://doi.org/10.3141/2290-02.
[36] J. Skibsted, R. Snellings, Reactivity of supplementary cementitious materials (SCMs) in
cement blends, Cem. Concr. Res. 124 (2019) 105799.
https://doi.org/10.1016/j.cemconres.2019.105799.
[37] S. Ramanathan, Reactivity of supplementary cementitious materials in model systems and
cementitious pastes, Ph.D. dissertation, University of Miami (2021).